The AEGK NRW project aims to use artificial intelligence to automatically identify green roofs and gravel roofs based on aerial imagery. The long-term goal is to establish a system for monitoring green roofs in North Rhine-Westphalia.

The goal of the project is to use aerial photographs and, where necessary, additional remote sensing, specialized, and geodata, as well as training data provided by LANUK, to identify green roofs and gravel roofs statewide using artificial intelligence. In the future, this method should be applicable on a regular basis to monitor the development of green roofs and to always have access to the most up-to-date information on which gravel roofs in the state are best suited for retrofitting with greenery.

In the planned project, current specialized methods will be tested for applicability and adapted to develop a reproducible and scalable methodology for clearly identifying green and gravel roofs statewide. The methodology can be developed in sub-areas or test regions. Ideally (depending on the required computing power and provided a sufficiently robust model is available), a dataset of green and gravel roofs covering the entire state of North Rhine-Westphalia (NRW) would be produced for use in the NRW Climate Atlas.